KR101947753B1 - Water pollution load evaluation apparatus and method using the l-thia acn-wq model - Google Patents

Abstract

The present invention relates to an apparatus and a method for water pollution load evaluation using an L-THIA ACN-WQ model, including a basic data input unit for input of basic data including at least one of soil information, hydraulic terrain information, and weather information including daily precipitation, daily average temperature, irradiation amount, and daylight hours by weather point, a curve number calculation unit calculating a curve number (CN) estimating effective rainfall based on the input basic data, a direct outflow calculation unit calculating a direct outflow from every hydrologic response unit (HRU) in a watershed to a stream based on the calculated curve number, a base outflow calculation unit determining infiltration by using an infiltration initial loss coefficient previously drawn along the curve number in the calculated direct outflow and calculating the amount of water flowing into an aquifer and a base outflow flowing to a stream based on the determined infiltration, and a pollution load calculation unit calculating a pollution load based on the calculated direct outflow and base outflow.

Description

BACKGROUND OF THE INVENTION Field of the Invention The present invention relates to a method and apparatus for evaluating wastewater pollution loading using an L-THIA ACN-WQ model,

The present invention relates to an apparatus and method for evaluating a watershed pollution load using an L-THIA ACN-WQ model, more specifically, a calculation of an initial loss amount estimation factor of penetration amount, application of a multiple weather point, application of a hydraulic repair structure, WATER POLLUTION LOAD EVALUATION APPARATUS AND METHOD USING THE L-THIA ACN-WQ MODEL BY USING L-THIA ACN-WQ MODEL AND MACHINE LEARNING .

In Korea, four major rivers are experiencing greenery every year, and the number of alerts issued is greatly increasing, and the public interest in the river environment is greatly increasing. Total nitrogen, total nitrogen (TN), total phosphorus (TP), nutrients such as water temperature, flow rate, water structure and nutrients are the most influential factors on algal growth. Total nitrogen (TN) and total phosphorus (TP) were calculated from non-point pollution sources such as some fertilizer components which were not consumed in total and fertilizer and livestock manure left in the vicinity of the farm, Are being introduced. In Korea, efforts to reduce total nitrogen (TN) flowing into rivers have been somewhat stagnant, but TP (total phosphorus) has undergone a lot of improvement efforts with the total amount system. Recently, increase of total nitrogen (TN) / total phosphorus Is becoming more prominent. Recently, the ratio of total nitrogen (TN) / total phosphorus (TP) concentration ratio in major rivers in Korea has been 23.8 ~ 36.6, which exceeds the ratio 16 that can affect the bird breeding. Therefore, total nitrogen (TN) and total phosphorus (TP) are very important to control green growth, and continuous monitoring for systematic management is an important point.

However, it is difficult to acquire continuous time - series water quality data of river / basin as the measurement data of water quality measurement network and gauging network. In addition, the CEM is mainly operated in the main streams of the four major rivers. Therefore, it is difficult to establish a river pollution load evaluation and management plan by using the water quality data of the water environmental monitoring network, which is a national monitoring network. Accordingly, in the management plan for the systematic management of the rivers, such as the 'Advanced apparatus for treating highly degradable organic matter and nutrients in the bottom and wastewater of Korean Patent Laid-Open Publication No. 2015-0018211 and a method for treating the wastewater using the same, Is used as an alternative. Among the various watershed models, the L-THIA model has been widely used to estimate rainfall-runoff and pollutant load in Korea as well as overseas because of relatively simple construction of input data and simple model improvement and operation. The L-THIA model is a version of a spreadsheet (Harbor, 1994) that estimates direct runoff-pollutant loads by estimating direct runoff using the CurveNumber (CN) method and considering the EventMean Concentration (EMC) , And L-THIA and L-THIA LID based on GIS have been supplemented and developed.

Recently, L-THIA ACN-WQ model was developed based on the monitoring data of land cover constructed through the environmental basic research project of the Ministry of Environment. The L-THIA ACN-WQ model estimates the runoff and pollutant loads through the 52 land-cover CN regressions and rainfall-grade EMC data for the land cover and soil, adds the underwater tracking module and links the QAUL2E model to the watershed / It is a model. However, the L-THIA ACN-WQ model has limitations in applying the watershed scale to the application of limitations on the calculation of specific CN section infiltration methods, the use of single meteorological stations, and the influence of downstream hydrographs on the installation of river repair structures. In addition, the L-THIA ACN-WQ model has various parameters and it is necessary to calculate the optimal parameters to reflect the environmental conditions of the watershed. However, the automatic correction tool for calculating the optimal parameters of the L-THIA ACN-WQ model Research has not been done.

SUMMARY OF THE INVENTION It is an object of the present invention to provide an apparatus and method for estimating wastewater pollutant load using an L-THIA ACN-WQ model that derives an initial loss amount coefficient of a permeation amount in direct runoff calculation .

The present invention also provides an apparatus and method for estimating wastewater pollutant load using an L-THIA ACN-WQ model for collecting meteorological data reflecting meteorological characteristics varying in time and space by region / watershed.

Further, the present invention provides an apparatus and method for estimating wastewater pollutant loading using an L-THIA ACN-WQ model that reflects the influence of hydraulic structures (dam, beam, and reservoir) .

It is another object of the present invention to provide an apparatus and method for estimating a wastewater pollutant load using an L-THIA ACN-WQ model that derives a representative optimal parameter or an optimal parameter for each wastewater area according to an environmental condition of the watershed.

In order to achieve the above object, an apparatus for evaluating a watershed pollution load using an L-THIA ACN-WQ model according to the present invention is provided with an apparatus and method for estimating watershed pollution load using weather data, A basic data input unit receiving basic data including at least one of soil information; A runoff curve index calculation unit for calculating a runoff curve (CN) that estimates the effective rainfall based on the input basic data; A direct runoff estimator that calculates a direct runoff from a hydrologic response unit (HRU) to a river based on the estimated runoff curve; Determination of the amount of infiltration using the infiltration initial loss coefficient derived based on the calculated outflow curve and the amount of inflow to the aquifer on the basis of the determined infiltration amount and the base flow rate to be flowed to the stream; And a pollutant load calculation unit for calculating a pollutant load based on the calculated direct outflow amount and the base runoff amount.

Also, the basic data input unit is characterized by receiving hydraulic terrain information for a digital elevation model (DEM) constructed through a resin topography map, soil information including land use map and hydrological soil group map information .

Also, the outflow curve index calculating unit calculates the outflow curve index using the precipitation amount of the meteorological information, the land use degree of the soil information, and the correction constant of the information of the hydrologic soil group.

Also, the base flow rate calculation unit may derive an initial loss amount coefficient based on a criterion that the initial loss amount calculated through the reduced outflow curve index should be larger than the outflow curve index before decreasing in the predetermined unit while decreasing the outflow curve index in a predetermined unit .

Also, the pollution load calculation unit calculates the pollutant load using the calculated direct flow amount and base flow amount, the entire watershed parameter derived in real time, the parameters by the geographical area, the total amount network flow rate and the water quality data, .

The parameter setting unit may further include a parameter setting unit for deriving and correcting parameters using basic data inputted in real time, wherein the parameter setting unit sets the first wastewater parameter of the first wastewater Parameter derivation; A second parameter derivation unit for deriving a parameter by geographical area through multiple calibration points in the watershed; And a parameter correcting unit for correcting the derived whole watershed parameter and the parameter for each owned area in real time.

In addition, the estimated pollutant loads can be plotted with a Load Duration Curve (LDC) or a standard load duration curve plotted through the daily flow rate and the target water quality, the daily flow rate and the actual water quality data And a pollutant load evaluating unit for estimating a pollutant load by plotting a QL Rating Curve (QLRC) comparing the Observed Load Duration Curve.

It also includes a lowland trace estimator that determines the flow rate in the stream based on the estimated direct runoff and base runoff.

In order to accomplish the above object, the present invention provides a method for estimating wastewater pollution loading using an L-THIA ACN-WQ model according to the present invention, wherein a baseline data input unit calculates a wastewater pollutant load based on a weather condition including daily precipitation, The method includes: receiving a basic data including at least one of information, hydraulic terrain information, and soil information; calculating an outflow curve index (CN) that estimates an effective rainfall based on input basic data, ; Estimating the direct runoff from all hydrologic response units (HRU) in the watershed to the river based on the runoff curve index calculated by the direct runoff estimator: The calculated runoff runoff Determining the amount of infiltration using the infiltration initial loss factor derived from the inflow curve index according to the outflow curve index, calculating the amount of water flowing into the aquifer and the amount of base flow flowing to the aquifer based on the determined infiltration amount; And calculating a pollutant load on the basis of the calculated direct outflow amount and the base outflow amount by the pollutant load calculation unit.

Also, the step of determining the amount of infiltration using the infiltration initial loss coefficient previously derived based on the calculated direct flow rate according to the outflow curve index, and calculating the amount of water flowing into the aquifer and the base flow rate flowing into the river based on the determined infiltration amount, The initial loss amount coefficient is derived based on a criterion that the initial loss amount estimated through the reduced outflow curve index should be larger than the outflow curve index before decreasing in the predetermined unit while decreasing the curve index in a predetermined unit.

Also, the step of calculating the pollutant load based on the estimated direct runoff amount and the base runoff amount is performed by using the calculated direct runoff amount and base runout amount, all the wastewater parameters derived in real time, the parameters of each wastewater zone, And the pollutant load is estimated using the expanded daily flow data.

In order to achieve the above object, an apparatus and method for estimating wastewater pollution loading using the L-THIA ACN-WQ model according to the present invention can be applied differently according to the curve number (CN) used in direct runoff calculation The results of this study can be used to estimate the adequacy of the estimation of the amount of infiltration by comparing deviations between the range of infiltration and the range of infiltration for the dominant soils such as sandy loam, clay loam and loam soil.

In addition, the present invention collects meteorological data reflecting meteorological characteristics that vary in time and space by region / watershed and collects meteorological data such as the maximum flow rate and the average flow rate according to the application of the single meteorological point and the multiple meteorological point, It is possible to comparatively analyze the maximum flow rate and the average flow rate according to the intervals, thereby reducing the uncertainty of the flow analysis result due to the rainfall due to the difference in the spatial abnormal characteristics.

In addition, the present invention reflects the influence of hydraulic structures (dams, beams, and reservoirs) that affect the sulfur characteristics of the stream, so that it is possible to easily grasp the stream flow fluctuation characteristics according to the hydraulic structure.

In addition, the present invention derives the representative optimal parameters of the entire watershed for the watershed terminal and derives the optimal parameters for each watershed through the performance of multiple monitoring points in the watershed to determine the environmental conditions (small or large scale) Therefore, it is possible to obtain an objective result by reducing the error between the simulated value and the measured value through the parameter correction.

In addition, the present invention optimizes parameters using machine learning to optimize L-THIA ACN-WQ model, and predicts flow rate and water quality in real time based on Web-GIS.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a diagram illustrating a configuration of an apparatus for evaluating watershed pollution load using an L-THIA ACN-WQ model according to the present invention. FIG.
FIG. 2 is a flowchart illustrating a procedure of a method for evaluating wastewater pollution loading using the L-THIA ACN-WQ model according to the present invention.
FIGS. 3 to 6 are diagrams for explaining the results of improvement of the method for estimating the amount of infiltration applied to the apparatus and method for estimating wastewater pollution loading using the L-THIA ACN-WQ model according to the present invention.
FIGS. 7 to 9 are views for explaining results of application of multiple virtual points applied to an apparatus and method for evaluating watershed pollution loading using the L-THIA ACN-WQ model according to the present invention.
FIGS. 10 and 11 are views for explaining the results of reflection of a river repair structure applied to an apparatus and method for estimating wastewater pollution loading using the L-THIA ACN-WQ model according to the present invention.
FIGS. 12 to 15 are diagrams for explaining the results of automatic parameter calibration applied to an apparatus and method for estimating wastewater pollution loading using the L-THIA ACN-WQ model according to the present invention.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail.

It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention. Like reference numerals are used for like elements in describing each drawing.

It is to be understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, . On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between.

The terminology used in this application is used only to describe a specific embodiment and is not intended to limit the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In the present application, the terms "comprises" or "having" and the like are used to specify that there is a feature, a number, a step, an operation, an element, a component or a combination thereof described in the specification, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. Hereinafter, the same reference numerals will be used for the same constituent elements in the drawings, and redundant explanations for the same constituent elements will be omitted.

First, the present invention combines the improved L-THIA ACN-WQ model with a pollution load evaluation system based on machine learning, and real-time daily flow rate and TN and TP are predicted through construction of L-THIA ACN-WQ system based on Web-GIS .

In addition, the study watershed of the present invention was selected as an aggregate watershed of the total amount of C, which is the upper basin of the Nakdong River basin (Kim et al., 2017, Song, 2017) Cave is the total unit of watershed of Sangju-Bo, which is located in the uppermost stream of Sangju-Bo, Nyeon-dongbo, Kumi-ba, Chilgokbo, The management area was selected as the area to be surveyed because it was expected that the management would continuously affect the growth and management of mainstream birds of 400.7 km. The upper part of class C is divided into the class A, Class B, Class A, Class B, Class Y, Class A, Class G, Class A, It has been reported that it has not been achieved. In addition, the water quality of the high turbid water has become a social problem due to the influence of the typhoon in 2002 and 2003, and in 2007 the water quality of Soo Yang Lake, Doam Lake, Gwangju Metropolitan City, It was designated as the first non-point pollution control area.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a diagram illustrating a configuration of an apparatus for evaluating watershed pollution load using an L-THIA ACN-WQ model according to the present invention. FIG.

1, the watershed pollution loading evaluation apparatus 100 using the L-THIA ACN-WQ model mainly comprises a basic data input unit 110, an outflow curve index calculation unit 120, a direct outflow calculation unit 130, A basin runoff calculation section 140, a bottom runoff calculation section 150, a parameter setting section 160, a pollutant load calculation section 170 and a pollutant load evaluation section 180.

The basic data input unit 110 receives basic data including at least one of daily precipitation, daily average temperature, solar radiation, weather information including the daylight time, hydraulic terrain information, and soil information for each weather spot.

The basic data input unit 110 includes a digital elevation model (DEM), a land use map, a hydrological map, and a hydrological model to take into account temporal changes in daily precipitation, daily average temperature, solar radiation, Soil groups also need data. The meteorological information was constructed from observational data of Taebaek, Bonghwa, Andong, Ueseong and Yeongdeok branches of meteorological stations of the Korea Meteorological Agency. DEM was constructed through the topographical map provided by the National Geography Center. Land use degree is constructed by the land cover map of the Ministry of Environment, and the hydrological soil group is constructed through the soil map provided by the RDA.

The outflow curve index calculation unit 120 calculates an outflow curve index (CN) that estimates the effective rainfall based on the input basic data.

The outflow curve index calculator 120 calculates the outflow curve index using the information of the precipitation amount of the weather information, the land use degree of the soil information, and the correction constant of the information of the hydrologic soil group.

[Equation 1]

Where CN is the CurveNumber, P is the precipitation (mm), and k is the calibration constant for the land use and hydrologic soil group.

The direct runoff calculation unit 130 calculates the direct runoff from the hydrologic response unit (HRU) in the watershed to the river based on the calculated runoff curve index. The direct discharge amount calculation formula is expressed by the following equations (2) and (3).

[Equation 2]

는 HRU별 발생되는 직접유출량(mm), P는 강수량(mm), S는 유역의 최대잠재보유수량(mm),

는 CN값을 의미함.here,

(Mm), P is the amount of precipitation (mm), S is the maximum potential quantity (mm) of the watershed,

Means the CN value.

[Equation 3]

는 하천으로 배출되는 최종 직접유출량(mm),

는 하루에 하천까지 도달되는 직접유출량(mm),

는 전날에 유달된 직접유출량(mm),

는 유달계수, TC는 time of concentration (hour)을 의미함.here,

(Mm) of the final direct discharge to the stream,

Is the direct discharge (mm) that reaches the stream per day,

(Mm), which was released on the previous day,

Is the supply coefficient, and TC is the time of concentration (hour).

Based on the determined infiltration amount, the amount of water flowing into the aquifer and the base flow rate flowing into the stream are calculated based on the infiltration amount using the infiltration initial loss coefficient previously calculated according to the calculated outflow curve index, .

Based on the criterion that the initial loss amount estimated through the reduced outflow curve index should be larger than the outflow curve index before decreasing in the predetermined unit while decreasing the outflow curve index in a predetermined unit, .

The descending section 150 determines the flow in the stream based on the estimated direct runoff and base runoff.

That is, the descending section 150 determines the flow rate in the stream by using the Muskingum method, which is used in the hydrology process, for the direct flow rate and the base flow rate flowing into the river

The parameter setting unit 160 derives and corrects parameters using basic data input in real time. The parameter setting unit 160 uses the method of deriving the optimal parameters for each sub-area through the flow rate and water quality measurement results of six points in the watershed and eight points in two dams, The parameters can be calibrated.

For this, the parameter setting unit 160 includes a first parameter derivation unit 161, a second parameter derivation unit 162, and a parameter correction unit 163.

The first parameter derivation unit 161 derives the entire watershed parameter of the entire watershed from the watershed end. All watershed parameters apply to small watersheds by applying the same parameters to all subwatersheds in the watershed.

The second parameter derivation unit (162) derives the own station-specific parameters via multiple correction points in the watershed. The parameters for each station are applied to large watersheds, which can be automatically calibrated step by step according to river flow from the upstream point to the downstream point.

The parameter correcting unit 163 corrects the entire watershed parameter and the parameter for each owning station derived in real time in each basic data input.

The parameter correcting unit 163 corrects the optimal parameters of the model by removing the outlier using the work flow expansion and the deep learning using the machine learning in consideration of the purpose of the model and the available correction data.

More specifically, the gauging network measures the flow rate and water quality on an average of 8 days intervals. The gauging interval of the gauging network has a limit to reflect the annual sulfur characteristics of the stream.

In particular, because the surveyor performs the monitoring directly, the safety of the surveyor is taken into consideration in the monitoring, and the measurement result of the concentrated rainfall or the rainy season, that is, the high flow rate is lacking. There is a problem that the uncertainty about the high flow rate can be greatly generated when the model parameters are corrected through the data of the gauging network. Accordingly, in the present invention, the daily flow expansion relation of the Nakdong River basin total point branch developed by utilizing Tenson Flow, a machine learning library disclosed by Google in an open source form, is applied to the end point C of the study watershed, The daily interval flow data was extended to the daily flow data and used as correction data.

And the most important factor in constructing a system based on Web-GIS is to maintain the accuracy of the model prediction. Especially, when the uncertainty in the prediction of the flow rate is increased, it is propagated to the prediction of the pollutant load to cause uncertainty. The L-THIA ACN-WQ 2 model according to the present invention develops an automatic correction module to reduce the uncertainty between the predicted value and the measured value, thereby deriving the optimum parameter. However, the presence of anomalies in the measured values has a great influence on the derivation of the optimal parameters. Deep learning technology was applied to remove the anomalies of the measured values. Deep Learning is an advanced artificial neural network model that uses algorithms to analyze and learn data, and to make predictions based on learned data. Deep Learning consists of several hidden layers between input and output, and complex nonlinear relationship modeling is performed between them. Although more complicated nonlinear relationship modeling is performed as the number of hidden layers increases, the computation time increases proportionally and it is important to construct a hidden layer for the purpose.

 Therefore, in the present invention, for the application of the removal of outliers for the deep learning, one flow volume observation is performed at the branch of the total amount measurement network of the Ministry of Environment, and a unit volume branch of the total volume B, which is the point where the one- Respectively. Through the Deep Learning algorithm of Google's Tensor Flow, the flow data of the Yangsan branch, which is the national land flow monitoring station at the B site, and the water level and the flow data of the upper stream basin of Dochun point were studied. The flow rate data were used to determine the abnormality of the measured values and to remove the abnormal values when applied as calibration data.

The pollutant load calculation unit 170 calculates the pollutant load based on the calculated direct outflow amount and the base runoff amount.

The pollutant load calculation unit 170 calculates the pollutant load (the amount of pollutant load) by using the calculated direct flow amount and base flow amount, the entire watershed parameter derived in real time, Real-time daily flow rate and TN and TP).

The pollutant load calculation unit 170 uses an average 8-day interval flow rate and water quality measurement data and an expanded daily flow data of the total amount measurement network. The flow rate and water quality data measured in the total flow measurement network during the one- As the interpolation concept. In other words, based on the actual flow rate and water quality, and the pollutant load value calculated through the calculation, only the flow rate for the measurement date is interpolated so that the pollutant load evaluation considering the entire sulfur characteristics is made.

The pollutant load evaluating unit 18 calculates the estimated pollutant load by a load duration curve (LDC) or a standard load duration curve drawn through the daily flow rate and the target water quality, The QL Rating Curve (QLRC), which compares the Observed Load Duration Curve plotted with the actual water quality data, is used to evaluate the pollutant load.

FIG. 2 is a flowchart illustrating a procedure of a method for evaluating wastewater pollution loading using the L-THIA ACN-WQ model according to the present invention.

Referring to FIG. 2, the watershed pollution load evaluation method using the L-THIA ACN-WQ model according to the present invention uses the above-described L-THIA ACN-WQ model watershed pollution load evaluation apparatus, A description thereof will be omitted.

First, basic data including at least one of weather information, hydraulic terrain information, and soil information including daily precipitation amount per day, daily average temperature, irradiation amount, daylight time, and the like is input (S100).

S100 is a digital elevation model (DEM) to take into account the weather and spatial distribution of daily precipitation, daily mean temperature, solar radiation, and daylight hours to consider temporal changes, land use, hydrological soil group data .

Next, an outflow curve index (CN) for estimating the effective rainfall is calculated based on the input basic data (S110).

In step S110, the outflow curve index is calculated using the precipitation amount of the meteorological information, the land use degree of the soil information, and the correction constant of the hydrologic soil group information.

Next, the direct runoff from the hydrologic response unit (HRU) in the watershed to the river is calculated (S120) based on the calculated runoff curve index.

Next, the infiltration amount is determined using the infiltration initial loss coefficient derived from the inflow curve index based on the calculated direct discharge amount, and the amount of water flowing into the aquifer based on the determined infiltration amount and the base flow rate flowing into the stream are calculated (S130).

In step S130, the initial loss coefficient is derived based on a criterion that the initial loss amount estimated through the reduced outflow curve index should be larger than the outflow curve index before decreasing in the predetermined unit while decreasing the outflow curve index in a predetermined unit.

Next, the flow rate in the river is determined based on the calculated direct flow rate and the base flow rate (S140).

Next, the pollution load calculation unit calculates the pollution load based on the calculated direct discharge amount and the base flow rate (S150).

In step S150, the pollutant load calculation unit calculates the pollutant load using the calculated direct flow rate and base flow rate, the entire watershed parameter derived in real time, the parameters by the geographical area, the gross mass flow rate data, and the expanded water flow data do.

In addition, parameters are derived and corrected using basic data input in real time before step S150 (S145).

Next, the estimated pollutant loads are plotted with a Load Duration Curve (LDC), or a standard load duration curve plotted through daily flow and target water quality is plotted through daily flow and actual water quality data The QL Rating Curve (QLRC), which compares the Observed Load Duration Curve, is evaluated to evaluate the pollutant load (S160).

FIGS. 3 to 6 are diagrams for explaining the results of improvement of the method for estimating the amount of infiltration applied to the apparatus and method for estimating wastewater pollution loading using the L-THIA ACN-WQ model according to the present invention.

Improved method of estimating the amount of permeation

The initial loss of L-THIA ACN-WQ model is calculated by assuming that the initial loss is the same as that of the direct runoff, and that the initial loss is high when the runoff curve is small Lt; / RTI >

Applying the initial loss factor of 0.2 when the outflow curve index is 99 and the initial loss factor of 0.2 when the outflow curve index is 98, the initial loss is 1.0367 and the initial loss is larger than CN 99. This is because the initial loss concept in the direct runout is applied as it is. When the runoff curve index is 98, the initial loss coefficient should be lowered to 0.098968, so that the initial loss is smaller than the runoff curve index 99, which is 0.5130. Therefore, in the present invention, the infiltration initial loss estimating coefficients according to the rainfall-runoff curve index are derived.

In order to verify the initial loss factor, we compute the infiltration rate using the initial loss factor and the infiltration rate using the Green-Ampt model.

The green-ampt model was used to estimate the amount of penetration of soil, clay loam, and loam soil with varying soil moisture content and suction head variation with rainfall of 5, 10, 30, and 50 mm. Soil moisture content was calculated by applying maximum - maximum, maximum - minimum, minimum - minimum porosity and effective porosity of each soil characteristics. And an improved penetration calculation method was used to estimate the penetration between 58 and 88 in the AMC-II standard runoff curve index of the field.

The amount of infiltration through Green-Ampt model was estimated to be 27.60m ~ 30.00㎜, average 28.66㎜, clay soil 3.45 ~ 12.46㎜, average 6.73㎜, loam 9.04 ~ 20.17㎜, average 12.14㎜, The penetration amount was 15.8mm on average. The AMC-II standard runoff curve index range of the field is known as 58 to 88. As shown in Fig. 3, the range of infiltration through the improved method of estimating the infiltration amount based on the rainfall amount of 30 mm was at least 10.91 mm to 22.34 mm, and the average infiltration amount of the field through the outflow curve method was 17.6 mm. Amount of infiltration of Green-Ampt model according to 6 conditions of 30mm of rainfall is 15.8mm on average. The infiltration amount of the field by the method of estimating infiltration based on the outflow curve index according to the present invention is 17.6mm, Of the respondents. The green-Ampt model and the improved method of estimating the amount of infiltration were found to be small, from about 0.8 mm at rainfall of 5 mm to about 0.9 mm at 10 mm and from 6.6 mm at 50 mm. In addition, as shown in FIG. 3, it is considered that the penetration amount according to the improved permeation amount calculation method is included in the range of 25 ppm to 75 ppm calculated from the Green-Ampt model.

Figure 4 shows the relationship between the initial loss factor and the initial loss factor, which can be applied differently according to the outflow curve index used in calculating the infiltration rate.

[Equation 4]

As shown in FIG. 5, the infiltration amount is calculated using the relational expression thus obtained, and the infiltration amount is generated even at a low rainfall amount at a low outflow curve index compared with the infiltration amount calculation result of the existing L-THIA ACN-WQ model. The increase in the amount of penetration is remarkable.

In the case of the existing L-THIA ACN-WQ model, the direct runoff at the rainfall of 50 mm was 0.33 mm and the infiltration was 12.62 mm when the runoff curve index was 55. This means that even if the evapotranspiration is 40% Uncertainty is generated. In the L-THIAACN-WQ model according to the present invention, the infiltration amount is 33.54 mm, and uncertainty about 4 mm is generated in the water balance analysis, and the uncertainty is greatly reduced compared to the existing L-THIA ACN-WQ model . This implies that flow simulations will have a significant effect on the flow rate simulation.

FIGS. 7 to 9 are views for explaining results of application of multiple virtual points applied to an apparatus and method for evaluating watershed pollution loading using the L-THIA ACN-WQ model according to the present invention.

Apply multiple virtual points

In the existing L-THIA ACN-WQ model, the meteorological data consists of four files that are independent of rainfall, temperature, sunshine hours and solar radiation. However, the weather data format of the L-THIA ACN-WQ model according to the present invention is independent The files consist of rainfall, temperature, sunshine hours, and solar radiation. Each independent file is assigned to each subwatershed and used as meteorological data input data for assigned weather files.

When the Jecheon, Jungsun, Wonju, and Yeongwol multi weather stations were applied to Pyeongchang A watershed and only Yeongwol single weather stations were applied to analyze the improvement effect by applying multiple weather stations, Koh Chang, Jeongeup, Buan multi weather And the flow characteristics of the buoyant single station were analyzed.

The maximum flow rate was 855.4㎥ / s, the average flow rate was 45.6㎥ / s, the maximum flow rate was 886.1㎥ / s, the average flow rate was 59.9㎥ / s, and the maximum flow rate The maximum flow rate was 108.8㎥ / s and the mean flow rate was 2.8㎥ / s during the simulation period in the case of a single meteorological station. / s, the maximum flow rate is 192.5㎥ / s, the average flow rate is 59.9㎥ / s, the maximum flow rate is about 30㎥ / s, and the average flow rate is about 4.8㎥ / s.

In the two watersheds, the flow simulation value was increased according to the application of the multiple meteorological stations, and the maximum flow rate and average flow rate were about twice or more difference .

The flow duration curve in Fig. 7 shows that the sulfur interval varies greatly in both Pyeongchang A and Koge A watershed according to the application of multiple weather stations. In Figs. 8 and 9, the FDC graph of the high-order A watershed shows a distinct variation characteristic compared to the Pyeongchang A watershed. In particular, a large difference is observed for the low-flow portion. When using a single meteorological point, the same rainfall is applied to all hydrologic response units (HRU) in the watershed for the same date, Because the flow rate is underestimated over the whole of the Mugang Wuxi basin due to the same day-to-day flow, the flow rate in the low-flow section is considered to be very small.

FIGS. 10 and 11 are views for explaining the results of reflection of a river repair structure applied to an apparatus and method for estimating wastewater pollution loading using the L-THIA ACN-WQ model according to the present invention.

River repair structure reflected

In order to reflect the influence of hydraulic structures such as dams, beams, and reservoirs affecting the sulfur characteristics of rivers, complex data such as water storage area, high water level, low water level, inflow volume, discharge volume, A simulation process is required. However, the existing L-THIA ACN-WQ model has many difficulties to reflect the characteristics of hydraulic structures such as dams, beams, and reservoirs. Therefore, in the L-THIA ACN-WQ model according to the present invention, the user inputs the discharge amount information so that the river repair structures such as dams and boats can be considered.

In the present invention, it is assumed that there is a dam at the end point of Juchun River in Pyeongchang A watershed and the applicability of the improved module is evaluated. The discharge amount was set to 6.25㎥ / s, which is the mean value of mid-range-flow interval belonging to the middle sulfur of FDC among the simulated inflows of the end subcatchment area of Juchun River.

As shown in FIG. 10, it is analyzed that the overall flow rate is increased at the low flow rate portion, and when the flow rate from October to November 2013 is observed, the flow rate of the river is greatly decreased according to the operation of the Juchun River Dam. This is because the influent flow rate from the Juchun River was controlled by the dam application and the flow rate of Pyeongchang River had a larger effect on the portion where the flow rate fluctuation before and after the dam application was larger.

11, the size of the sulfur section was increased, and the low-flow was 6.74 m / s slightly larger than 6.25 m 3 / s, which was set as the discharge amount, It is considered that the river flow fluctuation characteristics are properly improved and applied.

FIGS. 12 to 15 are diagrams for explaining the results of automatic parameter calibration applied to an apparatus and method for estimating wastewater pollution loading using the L-THIA ACN-WQ model according to the present invention.

Automatic calibration of parameters

In the present invention, a daily flow expansion module using machine learning was applied to select the optimal parameters reflecting the sulfur characteristics in the average data of 8 days interval and reflecting the whole sulfur characteristics. The application of the daily flow rate extension module is applied to the point C of the sample, and as a result of applying the expansion relation between January 2013 and December 2015, the calibration period is highly correlated with the result of the analysis of the sample C flow rate with R2 0.97 and NSE 0.96 . However, when the daily flow rate is expanded, the flow rate measured in the gauging network remains the same and the interpolation concept is applied only to the date of the flow error. This is because, due to the application of the expansion relation, there is a difference between the expanded flow rate and the measured flow rate value, which is propagated in the water quality correction, thereby increasing the uncertainty in the water quality correction.

And we developed an outlier removal module using Deep Learning to minimize uncertainty of optimal parameter selection according to actual data. The basic concept of the outlier removal using Deep Learning is that the flow rate at the point is correlated with the level and flow rate of the upstream watershed and the measured values outside the correlation are the outliers with artificial elements such as in- And does not use it to calibrate the model. Using the measurement results of Yangsham and Doshan daily flow observatories at Bukbap B, which is a target area, from July 2015 to June 2017 as learning data, the learning result between the river water level flow and the Yangsu flow rate was calculated as 8 It is applied to the daily flow rate data to judge the presence of an outlier.

Deep Learning is difficult to grasp and analyze the causes of outliers within the Deep Learning process because learning is done through complex nonlinear relations internally. Therefore, the trend analysis was conducted using data classified as outliers and other data. The data classified as outliers are in log form as shown in FIG. 13, and other data are in the form of polynomials. It is judged that input data is classified by input data through complex nonlinear relation in Deep Learning and it is decided whether there is an outlier according to the classification result.

As a result, 10 flow data, which is about 10% of 131 measured flow data, are discriminated as outliers from January 2013 to December 2015, which is the parameter correction period, and 121 actual flow data It was used for parameter calibration.

In more detail, in the present invention, the flow rate and water quality of the six spots A, B, C, A, G, A, and M, A excluding the discharge A and the half B, The automatic correction of the parameters of the L-THIA ACN-WQ model according to the present invention was performed through inflow information of the two points of Andong dam and Imha dam. In case A, the measured flow value is too small to be removed from the calibration point. As shown in Fig. 15, the average simulated values of eight points were simulated about 1.2 times larger than the actual values. However, seven points excluding the imha dam point , R2 0.59 ~ 0.86, and NSE 0.59 ~ 0.86 were satisfied with the efficiency range and confidence interval of the general model proposed by Moriasietal. (2007).

As described above, the apparatus and method for estimating wastewater pollution load using the L-THIA ACN-WQ model according to the present invention can estimate the initial loss amount of infiltration, which can be applied differently according to the curve curve (CurveNumber, CN) By calculating the coefficient, it is possible to evaluate the appropriateness of the estimation of the infiltration amount by comparing the deviation of the range of the infiltration amount and the range of the infiltration amount to the representative soil of the Korean field, the clayey soil, the clayey soil and the loam soil.

In addition, the present invention collects meteorological data reflecting meteorological characteristics that vary in time and space by region / watershed and collects meteorological data such as the maximum flow rate and the average flow rate according to the application of the single meteorological point and the multiple meteorological point, It is possible to reduce the uncertainty of the outflow analysis results due to the rainfall due to the difference of the spatial anomaly characteristics.

In addition, the present invention reflects the influence of hydraulic structures (dams, beams, and reservoirs) that affect the sulfur characteristics of the stream, so that it is possible to easily grasp the flow rate fluctuation characteristics according to the hydraulic structure.

In addition, the present invention derives the representative optimal parameters of the entire watershed for the watershed terminal and derives the optimal parameters for each watershed through the performance of multiple monitoring points in the watershed to determine the environmental conditions (small or large scale) Therefore, it is possible to obtain objective results by reducing the error between the simulated value and the measured value through the parameter correction.

In addition, the present invention optimizes parameters using machine learning to optimize L-THIA ACN-WQ model, and predicts flow rate and water quality in real time based on Web-GIS.

The functional operations described herein and the embodiments of the present subject matter may be implemented in digital electronic circuitry, computer software, firmware, or hardware, including any of the structures disclosed herein and their structural equivalents, It is possible.

Embodiments of the subject matter described herein may be implemented as one or more computer program products, that is, one or more of computer program instructions encoded on a type of program medium for execution by, or control of, a data processing apparatus May be implemented as a module. The type of program medium may be a propagated signal or a computer readable medium. A propagated signal is an artificially generated signal, such as a machine-generated electrical, optical or electromagnetic signal, generated to encode information for transmission to a suitable receiver device for execution by a computer. The computer-readable medium can be a machine-readable storage device, a machine-readable storage substrate, a memory device, a combination of materials that affect the machine readable propagation type signal, or a combination of one or more of the foregoing.

A computer program (also known as a program, software, software application, script or code) may be written in any form of programming language, including compiled or interpreted language, a priori or procedural language, Components, subroutines, or other units suitable for use in a computer environment.

The computer program does not necessarily correspond to the file of the file device. The program may be stored in a single file provided to the requested program or in a file that contains multiple interactive files (e.g., one or more of which store a module, subprogram, or portion of code) And may be stored in some (e.g., one or more scripts stored in a markup language document).

A computer program may be deployed to run on multiple computers or on one computer, located on a single site or distributed across multiple sites and interconnected by a communications network.

Additionally, the logic flows and structural block diagrams described in this patent document describe corresponding actions and / or specific methods supported by corresponding functions and steps supported by the disclosed structural means, It can also be used to build software structures and algorithms and their equivalents.

The processes and logic flows described herein may be performed by one or more programmable processors that execute one or more computer programs to perform functions by operating on input data and generating output.

Suitable processors for the execution of computer programs include, for example, any one or more of a general purpose and special purpose microprocessor and any type of digital computer. Generally, a processor will receive instructions and data from read-only memory, random access memory, or both.

A core element of a computer is a processor for performing one or more memory devices and instructions for storing instructions and data. In addition, the computer is generally operable to receive data from, or transmit data to, one or more mass storage devices for storing data, such as magnetic, magneto-optical disks, or optical disks, Or < / RTI > However, the computer need not have such a device.

The description sets forth the best mode of the invention, and is provided to illustrate the invention and to enable those skilled in the art to make and use the invention. The written description is not intended to limit the invention to the specific terminology presented.

Thus, while the present invention has been described in detail with reference to the above examples, those skilled in the art will be able to make adaptations, modifications, and variations on these examples without departing from the scope of the present invention. In other words, in order to achieve the intended effect of the present invention, all the functional blocks shown in the drawings are separately included or all the steps shown in the drawings are not necessarily followed in the order shown, It can be in the range.

100: Watershed Pollutant Load Assessment System Using L-THIA ACN-WQ Model.
110: basic data input unit
120: Runoff curve index
130: Direct spill and mountain government
140: Base flow rate calculation section
150: The government
160: Parameter setting unit
161: First parameter derivation unit
162: second parameter derivation unit
163: parameter correcting unit
170:
180: Pollution load evaluation unit

Claims (11)

A basic data input unit for receiving basic data including at least one of weather information including the daily precipitation amount per day, the daily average temperature, the solar radiation amount, the daylight time, the hydraulic terrain information, and the soil information;
A runoff curve index calculation unit for calculating a runoff curve (CN) that estimates the effective rainfall based on the input basic data;
A direct runoff estimator that calculates a direct runoff from a hydrologic response unit (HRU) to a river based on the estimated runoff curve;
Determination of the amount of infiltration using the infiltration initial loss coefficient derived based on the calculated outflow curve and the amount of inflow to the aquifer on the basis of the determined infiltration amount and the base flow rate to be flowed to the stream; And
A pollutant load calculation unit for calculating a pollutant load based on the estimated direct discharge amount and the base discharge amount;
Lt; / RTI >
Wherein the pre-derived infiltration initial loss coefficient is derived by substituting the outflow curve index into the following equation: L-THIA ACN-WQ model.

The method according to claim 1,
The basic data input unit receives the hydraulic terrain information for the digital elevation model (DEM) constructed through the resin topographic map, and the soil information including the land use degree and the hydrological soil group information. -THIA ACN-WQ Model for Watershed Pollutant Load Assessment. The method according to claim 1,
Wherein the outflow curve exponent is calculated by using an L-THIA ACN-WQ model, wherein the outflow curve index is calculated using the precipitation amount of the meteorological information, the land use degree of the soil information, and the correction constant of the hydrologic soil group information. Apparatus for evaluating pollution load. The method according to claim 1,
The base run-off amount calculation unit may derive an initial loss amount coefficient based on a criterion that an initial loss amount calculated through a reduced run-out curve index while decreasing a run-out curve index in a predetermined unit is greater than an run- WATER POLLUTION LOAD EVALUATION APPARATUS USING L-THIA ACN-WQ MODEL FEATURING. The method according to claim 1,
The pollutant load calculation unit calculates the pollutant load using the calculated direct flow rate and base flow rate, the entire watershed parameters derived in real time, the parameters by the geographical area, the total flow rate network flow rate and the water quality data, WATER POLLUTION LOAD EVALUATION APPARATUS USING L-THIA ACN-WQ MODEL FEATURING. The method according to claim 1,
Further comprising a parameter setting unit for deriving and correcting parameters using basic data input in real time,
Wherein the parameter setting unit comprises:
A first parameter derivation unit for deriving the entire watershed parameter of the entire watershed for the watershed terminal;
A second parameter derivation unit for deriving a parameter by geographical area through multiple calibration points in the watershed; And
A parameter correcting unit for correcting the derived whole watershed parameter and the own station-specific parameter in real time;
And the L-THIA ACN-WQ model. The method according to claim 1,
The calculated pollutant loads are plotted with a Load Duration Curve (LDC) or a standard load duration curve plotted through the daily flow rate and the target water quality, as well as daily flow and observed water quality data And a pollutant load evaluating unit for evaluating the pollutant load by plotting a QL Rating Curve (QLRC) comparing the Observed Load Duration Curve with the L-THIA ACN-WQ model Watershed pollution load assessment device. The method according to claim 1,
And an underflow trace calculation unit for determining a flow rate in the river based on the estimated direct runoff amount and the base runoff amount, using the L-THIA ACN-WQ model. Receiving basic data including at least one of weather information, hydraulic terrain information, and soil information including daily precipitation, daily average temperature, solar radiation, daylight time,
Estimating an outflow curve index (CN) for estimating an effective rainfall based on the inputted basic data by a flow curve index calculating section;
Estimating the direct runoff from the Hydrologic Response Unit (HRU) to the river based on the calculated runoff curve index by the direct runoff estimator:
Based on the determined infiltration amount, the amount of water flowing into the aquifer and the base flow rate flowing into the river are calculated based on the infiltration initial loss coefficient previously calculated according to the outflow curve index to the calculated direct outflow amount by the base flow outage calculation unit ; And
Calculating a pollutant load by the pollutant load calculation unit based on the calculated direct outflow amount and the base runoff amount;
/ RTI >
Wherein the initial loss amount coefficient of the inflow amount in the step of calculating the basin run-out amount is derived by substituting the outflow curve index into the following equation. ≪ tb >< TABLE >

10. The method of claim 9,
Determining the amount of infiltration using the infiltration initial loss factor derived from the calculated outflow curve according to the outflow curve index and calculating the amount of water flowing into the aquifer and the base flow rate flowing into the aquifer based on the determined infiltration amount,
Wherein the initial loss amount coefficient is derived based on a criterion that the initial loss amount estimated through the reduced outflow curve index should be larger than the outflow curve index before decreasing in the predetermined unit while decreasing the outflow curve index in a predetermined unit, Watershed pollution loading assessment method using THIA ACN-WQ model. 10. The method of claim 9,
The step of calculating the pollutant load based on the calculated direct outflow amount and the base outflow amount includes:
The estimated L-flow rate is calculated by using the calculated direct flow rate and base flow rate, the whole watershed parameters derived in real-time, the parameters for each station, the total flow network flow rate and the water quality data, Watershed pollution loading assessment method using THIA ACN-WQ model.

Images